Redundancy and Availability

Multiple modular units and duplicated auxiliaries let a site keep delivering firm power while any single unit or subsystem is serviced.

Availability by design

Firm power is only as good as the machine's uptime, so availability is engineered rather than hoped for. The main lever is modularity: a site made of several units can lose one to maintenance or fault and still deliver most of its capacity. This is the same logic that makes multi-engine and N+1 systems reliable across industry.

Layers of resilience

• Multiple units — one offline does not black out the site
• Redundant auxiliaries: pumps, power supplies, cryo capacity
• Linear geometry with open ends avoids disruptive plasma terminations
• Graceful fault response — plasma dumps into the expander, not onto in-vessel parts

Fault behavior helps

A tandem mirror has no large toroidal current to terminate, so it is not exposed to the violent disruptions that stress tokamaks. An off-normal event tends to end the burn by dumping plasma harmlessly out the open ends. That benign failure mode reduces the damage a fault does and shortens the return-to-service time, both of which raise availability.

Achieving high availability in practice depends on component lifetimes — magnets, collectors, first wall — and on maintainability. These are design-study targets validated progressively, starting with the test unit. The architecture is chosen to make high availability reachable, not to assert it as already proven.

The design intent is a plant that behaves like fleet infrastructure rather than a single fragile experiment: many units, duplicated auxiliaries, and benign faults combine so that scheduled and unscheduled outages of parts of the site rarely become outages of the whole.